Paper No.
09572
2009
AN OPEN SOURCE MECHANISTIC MODEL FOR CO2 / H2S CORROSION OF CARBON STEEL Srdjan Nešić(1), Hui Li, Jing Huang and Dusan Sormaz Institute for Corrosion and Multiphase Technology Department of Chemical Engineering Ohio University Athens, OH 45701
ABSTRACT A mechanistic model is developed to predict the corrosion rate caused by CO2, H2S, organic acids and/or O2. The aim of this model was to provide the corrosion community with a theoretically sound, simple and effective corrosion model for internal corrosion of mild steel lines. This model is available in the open literature and can be easily accessed on the internet. All the background information, including theories behind, data used for calibration, limitations, etc. is shared with users. In addition, the source code of the model, which has been written in object-oriented fashion, is open to the public to encourage utilization of any individual modules and development of add-on modules by third parties. The main features of this model include: prediction of uniform corrosion rate by CO2, H2S, organic acids and/or O2, simulation of iron carbonate film and iron sulfide film growth, identification of major corrosive species by quantifying respective contributions from various species, capability of distinguishing CO2/ H2S dominant corrosion processes, display and manipulation of polarization curves for CO2 dominant processes, or display of H2S concentration profile as a function of distance from steel surface for H2S dominant processes, capability of modifying the corrosion calculation process by adding user-defined reactions or excluding any undesired reactions.
Keywords: mechanistic model, corrosion prediction, CO2 corrosion, H2S corrosion, HAc corrosion
(1)
Corresponding author, E-mail address:
[email protected] (S. Nesic). Copyright ©2009 by NACE International. Requests for permission to publish this manuscript in any form, in part or in whole must be in writing to NACE International, Copyright Division, 1440 South creek Drive, Houston, Texas 777084. The material presented and the views expressed in this paper are solely those of the author(s) and are not necessarily endorsed by the Association. Printed in the U.S.A.
1
INTRODUCTION Carbon steel is extensively used in oil and gas pipelines due to the merit of low cost. These pipelines are frequently exposed to the aqueous environments containing aggressive species, such as CO2, H2S and organic acids which pose significant threat to the normal operation of pipelines. Intense research efforts have been made with the aim of understanding corrosion mechanisms caused by various corrosive species and factors that either promote or inhibit corrosion. As a result, a large number of publications have become available which deal with corrosion from different aspects. Although some issues remain unclear, most of the steps that are associated with CO2 and to a lesser degree H2S corrosion process are now understood. With available information, it is possible to establish a mechanistic model to predict corrosion rate and help further the understanding of corrosion process. In fact, a number of corrosion models for CO2 corrosion in wells and pipelines have been developed in recent years1. However, these models, most of which are proprietary and unavailable to public, often give a large scatter in the prediction due to different theories, assumptions and modeling strategies. This is aggravated by lack of transparency of the code behind the models. When more corrosive species are involved in the corrosion process, larger discrepancies are to be expected from various models. The project presented in this paper aimed at providing the corrosion community with a free mechanistic model for internal corrosion prediction of carbon steel pipelines. Strongly rooted in theories, this model can offer trustworthy predictions for a wide range of conditions that are typical for oil and gas pipelines, as will be shown in the following sections. This model, named FREECORP(2), was developed exclusively based on public information. All the information related to the model, including theories, assumptions, limitations and data used for calibration are shared with the users. To maximize the transparency of the model, the source code is also shared. With source code available, further development of the model can be easily achieved by adding new modules or by modifying existing modules. This provides a way to serve special needs that the users might have in their unique corrosion environments. This model is available in the open literature and can be easily accessed on the internet (http://www.corrosioncenter.ohiou.edu/freecorp). Being offered to the corrosion community, this model will hopefully:
Elevate the level of understanding and the prediction capability of mild steel corrosion as related to the oil and gas industry.
Ensure that best available science and technology is available to the corrosion engineers, implemented by using a transparent approach which is open for further development and improvement.
Increase the level of involvement of the broader corrosion community in developing better and more flexible tools fit for their intended purposes, an approach which will hopefully be mimicked in the future in other fields of corrosion.
Fulfill one of the key missions of the Institute for Corrosion and Multiphase Technology and Ohio University as a public institution: to educate the wider professional community and extend its reach beyond the pool of the current research sponsors in order to enable more effective dissemination of the already published knowledge and technology.
(2)
FREECORP is a product of the Institute for Corrosion and Multiphase Technology, Ohio University distributed under the GNU General Public License.
2
The model presented in this paper is capable of predicting uniform corrosion rate of carbon steel in the environment containing carbon dioxide, hydrogen sulfide, organic acid and/or oxygen. Respective contributions of various corrosive species are calculated which enables the identification of major corrosive species. Potentiodynamic sweep curve or H2S concentration profile is also shown depending on the dominating corrosion mechanism, which helps further elucidate the corrosion mechanism. The effect of iron carbonate growth is simulated using empirical relationships to enhance the accuracy of predictions. In the following sections, main theories of corrosion caused by carbon dioxide, hydrogen sulfide, organic acid and/or oxygen are reviewed, followed by implementation of the model. The model is then compared with experimental results under various conditions. In the final section, the limitations associated with the current version of the model are stated and the directions for future development are suggested.
THEORIES OF THE MODEL The theories and equations used in this model are attributed to various publications. Nonetheless, the main theories that serve as the basis of the present model are taken from three key papers2, 3, 4. To help understand the process described in the model, a brief review of the theories involved is given here. Detailed information about the corrosion process can be found in the original papers. Carbon Dioxide/ Acetic Acid Corrosion Carbon dioxide corrosion is a complicated process in which a number of chemical reactions, electrochemical reactions and transport processes occur simultaneously.2 Chemical reactions. Carbon dioxide is often present in the oil and gas pipelines. Various chemical reactions take place in the water phase due to the presence of carbon dioxide. These reactions have to be taken into consideration in order to get the accurate concentrations of corrosive species for further calculation. CO2 is soluble in water: CO 2 ( g ) ⇔ CO 2 ( w )
(1)
Hydration of aqueous CO2 produces a weak acid, known as carbonic acid, H2CO3.: CO 2 ( w ) + H 2 O ⇔ H 2 CO3
(2)
Carbonic acid can then partially dissociate in two steps to form bicarbonate and carbonate ions: H 2 CO3 ⇔ H + + HCO3−
(3)
HCO3− ⇔ H + + CO32−
(4)
Homogenous dissociation reactions (3) and (4) proceed much faster than the other simultaneously occurring processes in the system. CO2 dissolution reaction (1) and particularly CO2 hydration reaction
3
(2) have been known to be much slower (rate controlling) and could lead to local non-equilibrium in the solution. In some environments organic acids, particularly low molecular weight organic acids, are found primarily in water phase and can lead to corrosion of mild steel. Acetic acid, (HAc) which is the most prevalent type of organic acids found in brines, can be treated as the representative of other types of organic acids as similar corrosiveness was found for all types of low molecular weight organic acids. HAc is a weak acid which is rather volatile – a property that makes it a major concern in top-of-line corrosion (TLC). It partially dissociates in the water phase to give away H+ and Ac- , as indicated by reaction (5). However it is stronger than H2CO3, and therefore serves as the main source of H+ when similar concentration of the two acids are encountered. (5)
HAc ⇔ H + + Ac −
In the temperature range of interest (20 – 100oC), HAc is mainly found in the aqueous phase and therefore corrosion caused by HAc is not strongly affected by presence of HAc vapors (other than in the case of TLC). This is different from CO2 corrosion where gaseous CO2 controls the amount of H2CO3 in the aqueous phase and therefore determines the rate of CO2 corrosion. Another important chemical reaction that needs special attention is the formation of iron carbonate in the circumstance where concentrations of ferrous ion and carbonate ion exceed the solubility limit of iron carbonate, as indicated by reaction (6): (6)
Fe 2+ + CO32− ⇔ FeCO3( s )
This reaction is a heterogeneous reaction as nucleation of solid iron carbonate occurs preferably on the steel surface or inside the pores of the present film. The precipitation of FeCO3(s) often plays a significant role in the corrosion process because the FeCO3(s) layer may increase the mass transfer resistance for the corrosive species as well as reduce the available steel surface exposed to the corrosive solution. In fact, in many cases, the CO2 corrosion rate is largely controlled by the presence of the FeCO3(s) layer. Electrochemical reactions. For CO2 corrosion, several electrochemical reactions have been identified as contributors to overall corrosion rate of mild steel. Dissolution of iron is the dominant anodic reaction. This reaction proceeds via a multi-step mechanism which is mildly affected by pH and CO2 concentration. In the range of typical CO2 corrosion, e.g. pH >4, the dependency of pH tends to diminish. Therefore, for practical purposes, this reaction can be treated as pH independent for CO2 corrosion. At the corrosion potential (and up to 200 mV above), this reaction is under charge transfer control and therefore the electrochemical behavior can be described using Tafel’s law. (7)
Fe → Fe 2+ + 2e
Hydrogen ion, H+, reduction is one of the main cathodic processes:
4
(8)
2 H + + 2e − → H 2
This reaction is limited by how fast the H+ can be transported from the bulk solution to the steel surface through the mass transfer layer (including liquid boundary layer and the FeCO3(s) layer if it exists). Higher corrosion rates are often experimentally observed for CO2 system comparing to strong acid solution (such as HCl) at the same pH. However, for practical CO2 system where pH>4, the limiting flux of this reaction would be small due to the relative low concentration of H+. This suggests that CO2 also plays certain role in H+ reduction. This can be explained by the fact that the homogeneous dissociation of H2CO3 provides an additional reservoir for H+ ions, which then absorbs on the steel surface and gets reduced according to reaction (8). Apparently, any rapid consumption of H+ can be readily replenished by reactions (3) and (4). Thus for pH>4 the presence of CO2 leads to a much higher corrosion rate than would be found in a solution of a strong acid at the same pH. In the vicinity of the steel surface, another electrochemical reaction can take place as well: H2CO3 adsorbs at the steel surface and is directly reduced according to reaction (9). This is referred to as “direct reduction of carbonic acid”5. In fact, this reaction is just an alternative pathway for the same cathodic reaction - hydrogen evolution as addition of reaction (3) and (8) leads to reaction (9). The distinction is only in the pathway, i.e in the sequence of reactions. (9)
2 H 2CO3 + 2e − → H 2 + 2 HCO3−
The rate of this additional hydrogen evolution due to the presence of CO2 is mainly controlled by the slow CO2 hydration step (2) and is a strong function of H2CO3 concentration which directly depends on partial pressure of CO2. Although oxygen is not a common corrosive specie in oil and gas pipeline systems, it can invade the system by inappropriate operation or incomplete de-oxigenation of water chemical solutions injected into the system. Oxygen can contribute to corrosion process by oxygen reduction reaction which is limited by transport of oxygen through the mass transfer layer, as shown below: (10)
O2 + 2 H 2 O + 4e − → 4OH −
Acetic acid is known to be one of the species that attacks mild steel. Studies have shown that it is the undissociated (“free”) HAc and not the acetate ion Ac- that is responsible for corrosion. The presence of organic acids is a major corrosion concern particularly at lower pH and high temperature as more HAc would be generated under these conditions according to reaction (5). Being a weak i.e. partially dissociated acid, HAc provides an additional source of H+, which then adsorbs at the steel surface and reduces according to the cathodic reaction (8). Following the reasoning with H2CO3, it is also possible that the HAc molecule itself is adsorbed at the steel surface and gets reduced, which is often referred to as the “direct HAc reduction” pathway: (11)
2 HAc + 2e − → H 2 + 2 Ac − Another possible pathway for hydrogen evolution is direct reduction of water:
(12)
2 H 2O + 2e − → H 2 + 2OH −
5
Compared to the cathodic reactions described above, this pathway is very slow and often can be neglected in practical CO2 corrosion environments. However, under peculiar conditions, such as very low partial pressure of CO2 (PCO2 6), this reaction may become significant and contribute to the overall corrosion process. Therefore, even if rarely relevant, this reaction has been included in the model for completeness. Transport. Due to electrochemical reactions, certain species are produced or consumed on the steel surface. Concentration gradients are therefore established between bulk solution and steel surface, which leads to molecular diffusion. In addition, pipelines are often subject to turbulent flow. Turbulent eddies can usually penetrate into the boundary layer and greatly enhance the transport of species. Compared to some fast electrochemical reactions, such as hydrogen evolution (8), mass transfer of H+ proceeds much slower and therefore the rates of the overall reaction will be limited by transport, i.e. how fast the species can move through the mass transfer layer and any solid corrosion product layer. It is therefore essential to incorporate transport phenomenon into the model in order to give an accurate description of the overall corrosion process. The effect of mass transport can be readily captured by using the concept of mass transfer coefficients. Numerical calculation for CO2/HAc corrosion CO2 and HAc corrosion are both electrochemical process. The corrosion rate can therefore be calculated based on total current density of the anodic reaction (or the sum of cathodic reactions). CR =
(13)
ia M w, Fe
ρ Fe nF
Where, CR: corrosion rate, unit conversion factors are needed for appropriate unit; ia: anodic current density, A/m2; Mw,Fe: atomic mass of iron, kg/mol; ρ Fe : density of iron, kg/m3; n: number of moles of electrons involved in iron oxidation, 2 mole/mol; F: Faraday’s constant. The anodic current density of iron oxidation is obtained by, (14)
E corr − E rev , Fe
ia , Fe = io , Fe × 10 Where,
b Fe
ia,Fe: current density of iron oxidation, A/m2; io,Fe: exchange current density of iron oxidation, A/m2; E: corrosion potential, V; Erev,Fe: reversible potential of iron oxidation, V; bFe: Tafel slope of iron oxidation, V.
The current density of individual cathodic reaction is given by,
6
(15)
1 1 1 = + ic ict ilim
Where,
ic : current density of any cathodic reaction,A/m2; ict : component of charge transfer current density, A/m2;
ilim : component of limiting current density , A/m2; The charge transfer current density of cathodic reactions can be calculated by,
ict = io ×10 Where,
Erev − Ecorr b
(16)
⋅η FeCO3 ⋅η FeS
iO : exchange current density of cathodic reactions, A/m2; E rev : reversible potential of cathodic reactions, V; b : Tafel slope of cathodic reactions, V;
η FeCO ,η FeS : scale factors due to formation of FeCO3 and FeS film, respectively. 3
Most of cathodic reactions are mass transfer limited. The limiting current density can be calculated as, d ilim = η FeCO3 ⋅ η FeS ⋅ k m Fc j
(17)
Where, k m : mass transfer coefficient of corrosive species, m/s; c j : bulk concentration of corrosive species, mol/m3; However, carbonic acid reduction is limited by slow reaction rate of CO2 hydration, as shown in reaction 2. The limiting current density of this reaction is calculated in a different fashion, as shown below,
(
r f ilim, H 2CO3 = FcCO2 η FeCO3 ⋅ η FeS ⋅ D H 2CO3 K hyd k hyd
Where,
)
0.5
f
(18)
cCO2 : concentration of CO2 in bulk solution, mol/m3; D H 2CO3 : diffusion coefficient of H2CO3 in water, m2/s; k m , H 2CO3 : mass transfer coefficient of H2CO3, m/s;
Khyd: equilibrium constant for CO2 hydration reaction; d k hyd : forward reaction rate constant for CO2 hydration reaction; f: flow factor affecting CO2 hydration, which is given by, f =
(19)
1 + e −2δ m / δ r 1 − e − 2δ m / δ r
7
Where δ m and δ r are defined as mass transfer and reaction layer thickness, respectively, and can be calculated as,
δm =
δr =
(20)
DH 2CO3 k m , H 2CO3
(21)
DH 2CO3 K hyd f k hyd
The unknown corrosion potential, Ecorr, can be found based on charge balance on the steel surface. na
nc
1
1
(22)
∑ i a = ∑ ic Where,
ia, ic: anodic and cathodic current density, respectively, A/m2; na, nc: total numbers of anodic and cathodic reactions, respectively.
Once the corrosion potential is found, the corrosion current density can be calculated by using equation (14) as iron oxidation is normally the only one anodic reaction involved in the corrosion process even though the number of cathodic reactions is variable depending on the corrosive environment. Hydrogen Sulfide Corrosion In contrast with CO2 corrosion, which is electrochemical in nature, H2S corrosion of mild steel is considered to proceed predominantly via a solid state reaction: Fe + H 2 S ⇔ FeS (s ) + H 2
(23)
The term solid state reaction refers to the fact that both initial and final state of iron are solid. A plausible mechanism for H2S corrosion of mild steel proposed by Sun et. al.4 is adopted in this model. According to this theory, H2S initially absorbed onto the steel surface reacts very fast with iron to form a very thin (
